What independent datasets (TRAC, academic studies) say about interior ICE arrests and deportations in 2025–2026?
Executive summary
Independent datasets and academic analyses indicate a sharp increase in interior ICE arrests, detentions, and removals across 2025 with evidence that much of the enforcement surge was carried out in U.S. communities rather than solely at the border [1] [2]. Key independent sources — the Deportation Data Project and TRAC — show larger detention populations, rising numbers of deportations in FY2025–FY2026, and substantial data gaps and inconsistencies that complicate precise counts [3] [4] [5].
1. What the independent datasets actually measure and how they were produced
The Deportation Data Project (public FOIA-derived ICE extracts) posts individual-level arrest, detainer and detention records through October 15, 2025 and provides processed tables and documentation to enable analysis, while warning of missing fields and table omissions such as removals/encounters in some releases [1] [3] [5]. TRAC at Syracuse uses ICE-released statistics and its own aggregations to produce removals and detention tallies; TRAC reported 56,392 ICE removals so far in FY2026 and has tracked detention population changes through November 2025 [4] [6]. Academic and policy shops (Migration Policy, Brookings, Prison Policy Initiative) combine these public datasets with ICE detention-management reports to estimate broader removal totals and characterize enforcement shifts [2] [7] [8].
2. Core findings: rising interior enforcement, detention, and removals
Multiple independent analyses find that interior enforcement rose sharply in 2025: Migration Policy reports more deportations originating from interior operations in FY2025 than Border Patrol apprehensions for the first time in recent history, and that ICE detention populations grew toward record levels during 2025 [2]. TRAC and other trackers show ICE detention averages near the tens of thousands (TRAC: about 65,000 detained as of late November 2025) and report that a large share of detainees lacked criminal convictions — roughly three quarters in TRAC’s November snapshots [6]. Brookings’ modeling, drawing on Deportation Data Project and ICE reports, estimated roughly 310,000–315,000 removals for 2025, reflecting both border and interior removals but acknowledging gaps [7].
3. Who was being arrested and where — nuance in the demographic and criminal-history picture
Independent datasets highlight that many people arrested in the interior had no serious criminal convictions: TRAC and Prison Policy note a high share of detainees and arrestees without convictions or with only minor offenses such as traffic charges, and state-level analyses (e.g., New Jersey vs. New York comparisons) show variation in criminal-history profiles across field offices [6] [8]. Local reporting from Honolulu Civil Beat shows that U.S. island jurisdictions experienced spikes consistent with national trends, and that arrests appeared increasingly likely at ICE offices and civil courthouses [9].
4. Data limitations, discrepancies, and why exact tallies remain contested
Independent trackers repeatedly warn that ICE releases are incomplete and inconsistent: Deportation Data Project notes missing identifiers and shifted coding that block tracing individuals across tables, and in some releases ICE withheld complete removals/encounters tables citing data errors [5] [3]. Migration Policy and Brookings stress selective public reporting by ICE and reliance on piecing together detention-management reports plus FOIA datasets, which produces uncertainty in month-to-month totals and forces model-based estimates [2] [7]. Prison Policy underscores location and identifier gaps that prevented matching high-profile claims (like a Chicago “plane” of detainees) to the dataset [8].
5. Interpretation, competing narratives, and what the datasets do and do not prove
Taken together, independent datasets and academic estimates provide strong convergent evidence that interior ICE arrests, detentions, and deportations increased substantially in 2025 and into FY2026, and that enforcement shifted toward interior operations rather than border-only activity [2] [4] [7]. However, these sources also document substantial uncertainty about precise counts, about how many removals were interior-initiated versus border-initiated, and about the granular identities of arrested persons because of missing fields and shifting ICE reporting practices [3] [5]. Alternative viewpoints from ICE’s public materials stress categories (criminal convictions, pending charges, immigration violators) but independent processors warn that those categories are sometimes inconsistently applied across releases, complicating direct comparison [10] [5].